A cross-over study investigating specific aspects of neuropsychological performance in hyperbaric environments

(1)

A cross-over study investigating specific aspects of

neuropsychological performance in hyperbaric environments

Charles Halloran van Wijk

Thesis presented in fulfilment of the requirements for the degree of Master of Science in the Faculty of Medicine and Health Sciences

at Stellenbosch University

Supervisor: Dr Willem Albertus Jacobus Meintjes

(2)

DECLARATION

By submitting this thesis electronically, I declare that the entirety of the work contained therein is my own, original work, that I am the sole author thereof (save to the extent explicitly otherwise stated), that reproduction and publication thereof by Stellenbosch University will not infringe any third party rights and that I have not previously in its entirety or in part submitted it for obtaining any qualification.

April 2014

(3)

ABSTRACT

The commercial and military deep diving environment is typically a low visibility environment, where dependence on the visual senses often needs to be replaced by a reliance on tactile senses.

This thesis reviewed the current knowledge regarding neuropsychological manifestations of nitrogen narcosis and exposed a number of shortcomings in the current body of knowledge. In particular, the human performance effects of hyperbaric exposure on tactile perception and memory have not been systematically studied. It is further not clear, how exactly psychological factors (e.g. anxiety, mood states) and biographical factors (e.g. age, education, technical exposure, experience) might influence tactile perception and memory performance under conditions of hyperbaric exposure. The correlation between subjective experiences of narcosis, tactile performance, and psychological and biographical variables is also unknown. This study thus set out to investigate certain neuropsychological aspects of nitrogen narcosis, with special reference to tactile perception and memory, and to examine the relationships of tactile performance with other psychological and biographical factors.

The effects of experimental hyperbaric exposure (EHE) on tactile (form) perception and tactile shape memory were examined by testing these functions at 6 ATA and 1 ATA, using a cross-over design where two groups completed the same tasks, in opposite sequence. The psychological variables included trait anxiety, transient mood states, and subjective ratings of narcosis, while the biographical variables included age, education, and previous technical exposure.

The results demonstrated the detrimental effect of nitrogen narcosis on tactile form perception and manipulation, irrespective of the sequence of testing. It also demonstrated this effect on tactile form memory, although the sequence of testing also played a role here.

Higher trait anxiety was associated with poorer recall, and tension was associated with a larger decrement in recall performance, while fatigue was associated with poorer task completion. Subjective experiences also played a role, where feelings of physical anxiety (i.e. increased arousal) were associated with better recall, and feelings of cognitive suppression (decreased arousal) were associated with a larger decrement in recall performance. Lower academic attainment was associated with poorer recall, while higher diving qualification was associated with better recall. Performance on the surface was a good predictor of performance at depth. Qualitative analysis rendered three themes, namely focus vs. distraction, following instructions, and shape memory. Psychometric properties of the subjective narcosis measure were also reported.

(4)

Theoretical implications include support for the slowed information processing model when completing complex neuropsychological tasks, as well as support for the memory model, thus suggesting that this particular pattern of memory impairment occurs because encoding under narcosis produces a weaker memory trace than normal.

Lastly, the study has a number of implications for industry. For example, divers need to compensate for slowed task completion by, firstly, planning more time to complete complex tasks, and secondly, by practicing those tasks prior to the actual deep dive (either on the surface or in shallow water). The need for using additional forms of recording of events or objects at depth, to aid memory encoding and subsequent recall at surface was also emphasised.

(5)

OPSOMMING

Kommersieële en militêre duik vind dikwels plaas in ‘n omgewing met swak sig, waar duikers moet staatmaak op taktiele sintuie, eerder as op visuele sintuie.

Die tesis begin met ‘n oorsig oor die huidige kennis rakende neurosielkundige verskynsels van stikstof narkose, en het ‘n aantal tekortkominge gevind. Meer spesifiek, die menslike faktor in die effek van hiperbariese druk op taktiese persepsie en geheue is nog nie sistematies bestudeer nie. Dit is verder nie duidelik presies hoe sielkundige faktore (angs, gemoedstoestande) en demografiese faktore (ouderdom, opvoeding, tegniese blootstelling, ondervinding) taktiele persepsie en geheue onder toestande van hiperbariese druk sou beïnvloed nie. The korrelasie tussen die subjektiewe ervaring van narkose, taktiele taakverigting, en sielkundige en biografiese veranderlikes is ook nie bekend nie. Die studie het verskeie neurosielkundige aspekte van stikstof narkose, met spesifieke verwysing na taktiele persepsie en geheue, sowel as die verhouding tussen taktiele prestasie en sielkundige en biografiese faktore ondersoek.

Die effek van hiperbariese druk op taktiele persepsie en geheue is ondersoek deur hierdie funksies te toets by 6 en 1 ATA, deur middel van ‘n oorkruis studie ontwerp, waar twee groepe die take voltooi het, in teenoorgestelde volgorde. Die sielkundige veranderlikes het bestaan uit trek-angs, tydelike gemoedstoestande, en die subjektiewe evaluering van narkose, terwyl die biografiese veranderlikes ouderdom, opvoeding, en vorige tegniese blootstelling ingesluit het.

Die resultate het die nadelige effek van stikstof narkose op taktiele vorm persepsie en manipulasie gedemonstreer, ongeag die rigting van toetsing. Dit het ook hierdie effek op taktiele vorm geheue gedemonstreer, hoewel die rigting van toetsing wel hier ‘n rol gespeel het.

Hoër trek-angs was geassosieër met swakker herroeping, en spanning met ‘n groter agteruitgang in herroeping, terwyl matheid geassosieer was met swakker taakvoltooiing. Subjektiewe ervarings het ook ‘n rol gespeel, met ervarings van fisiese spanning (verhoogde opwekking) geassosieer met beter herroeping, en ervarings van kognitiewe onderdrukking (verlaagde opwekking) met groter agteruitgang in herroeping. Laer akademiese kwalifikasie was geassosieer met swakker herroeping, terwyl hoër duik kwalifikasie geassosieer was met beter herroeping. Taakverrigting op die oppervlak was ‘n goeie voorspeller van prestasie op diepte. Kwalitatiewe analiese het drie temas geidentifiseer, naamlik fokus vs. afleibaarheid, die volg van instruksies, en vorm geheue. Die psigometriese eienskappe van die subjektiewe narkose meetinstrument is ook gerapporteer.

(6)

Teoretiese implikasies van die studie sluit in ondersteuning vir die vertraagde prosesseringsmodel, wanneer komplekse neurosielkundige take voltooi word, sowel as ondersteuning vir die model vir hierdie spesifieke herroepingspatroon wat ‘n swakker geheuespoor laat wanneer enkodering plaasvind onder toestande van narkose.

Die studie het ook praktiese implikasies vir industrie. Dit is byvoorbeeld nodig om te kompenseer vir vertraagde taakvoltooïng deur, eerstens, die beplanning vir meer tyd om komplekse take te voltooi, en tweedens, deur daardie take te oefen voor die diep duik plaasvind. Die noodsaaklikheid vir additionele maniere om gebeure of voorwerpe op diepte vas te lê is ook beklemtoon.

(7)

LIST OF FIGURES

Figure 2.1: Effects of exposure to compressed air (Fowler et al., 1985). ... 6

Figure 4.1: Cross-over design groups. ... 34

Figure 5.1: BRUMS profile of total group according to mood states. ... 49

Figure 5.2: Means of TNT-tt performance under different conditions and in different sequence. ... 51

Figure 5.3: Means of TNT-dr performance under different conditions and in different sequence. ... 52

Figure 5.4: Scatterplot of TNT-tt at 1 and 6 ATA. ... 55

(11)

LIST OF TABLES

Table 2.1: Effects of exposure to compressed air at increased pressure and depth. ... 12

Table 2.2: Studies investigating choice reaction time. ... 14

Table 2.3: Studies investigating visual scanning and tracking. ... 14

Table 2.4: Studies investigating time estimation. ... 14

Table 2.5: Some studies investigating verbal memory. ... 16

Table 2.6: Studies investigating reasoning. ... 18

Table 2.7: Studies investigating fine visual-motor coordination. ... 20

Table 5.1: Distribution of academic achievement and current diving courses. ... 48

Table 5.2: Mean scores and variance of BRUMS mood states. ... 50

Table 5.3: Results from repeated measures ANOVA. ... 50

Table 5.4: Correlations between psychological variables and neuropsychological performance. ... 53

Table 5.5: Correlations between SHAS scores and neuropsychological performance (N=40). ... 53

Table 5.6: Correlations between biographical variables and neuropsychological performance. ... 54

(12)

LIST OF ACRONYMS

ATA atmosphere absolute BQ Biographical Questionnaire BRUMS Brunel Mood State

CNS central nervous system

EHE experimental hyperbaric exposure fsw feet sea water (unit)

GABA gamma-amino butyric acid

GP Grooved Pegboard

HPNS High pressure nervous syndrome IGN inert gas narcosis

kPa kilopascal (unit)

LTM long-term memory

MRA Multiple Regression Analysis

msw metres sea water

NN nitrogen narcosis

PICF Participant Information and Consent Form POMS Profile of Mood Scales

SHAS Subjective High Assessment Scale STAI State-Trait Anxiety Inventory

STPI State-Trait Personality Inventory, Trait Anxiety TMD total mood distress

TNT Tupperware Neuropsychological Task WAIS-R Wechsler Adult Intelligence Scale – Revised WMS-R Wechsler Memory Scale – Revised

(13)

CHAPTER 1: INTRODUCTION

The effects of deep diving on human performance have been subject to serious scientific investigation for almost two centuries. The effects of working underwater and in greater depth, and with more technologically advanced equipment and material, is increasingly better understood. However, within this field of study, human performance through the tactile modality is a neglected area of research.

The commercial and military deep diving environment is typically a low visibility environment, where dependence on visual senses often needs to be replaced by a reliance on tactile senses. The purpose of this study is to investigate certain neuropsychological aspects of nitrogen narcosis (NN), with special reference to tactile perception and memory, and to examine the relationships of tactile performance with other psychological and biographical factors.

The effects of experimental hyperbaric exposure (EHE) on tactile performance (referring to form perception, mental manipulation and fine motor coordination), and tactile memory (referring to shape memory) were examined through testing of these functions at 6 ATA1 and 1 ATA, using a cross-over design.

The psychological variables include trait anxiety, transient mood states, and subjective ratings of narcosis, while the biographical variables include age, education, and previous technical exposure.

Chapter 2 will present a review of the available literature, which will culminate in a formulation of the specific objectives of the study in Chapter 3. Chapter 4 will describe the methodologies employed to achieve these objectives. The findings will be reported in Chapter 5, and Chapter 6 will discuss these results, before concluding with some of the lessons that can be learnt from this study.

1

(14)

CHAPTER 2: LITERATURE REVIEW

2.1. Introduction

This chapter will review the available literature on neuropsychological performance in hyperbaric environments. The first section (Section 2.2) will provide an overview of inert gas narcosis (IGN), with specific reference to the effects of NN on human behaviour. As this study is primarily interested in tactile functions, e.g. in form perception, fine motor manipulation of shapes, and tactile shape memory, specific emphasis will be placed on understanding the underlying psychological principles that regulate such information processing. A number of other factors – both in terms of personal and environmental variables – are further implicated in affecting human performance under water and at depth, and these will be reviewed in Section 2.3. Section 2.4 will briefly review a number of relevant psychological constructs, with specific reference to anxiety, while Section 2.5 will introduce tactile perception. Section 2.6 will summarise the salient findings, to be followed by the problem statement in Section 2.7.

2.2. Inert gas narcosis

2.2.1 Definition and description

Inert gas narcosis (IGN) refers to the effects of breathing inert gases at pressures higher than 1 atmosphere absolute (ATA) (Bennett, 1993). In the broadest sense, the term ‘narcosis’ refers to the reversible depression of function of an organism (Fowler, Ackles & Porlier, 1985). Nitrogen is the most common inert gas found in air, and narcosis is most commonly found in compressed air diving. Nitrogen narcosis thus describes a group of symptoms characterised by deteriorating cognitive and neuromuscular functions, and disturbed mood and behaviour (Petri, 2003). For this reason, NN may have a negative impact on underwater work performance and can be a factor in dive-related accidents (cf. Kneller, Higham & Hobbs, 2012).

The threshold for the occurrence of nitrogen narcosis is usually regarded to be at a depth of about 30 metres of seawater (4 ATA) during compressed air breathing (Bennett, 2004), with the signs and symptoms becoming increasingly pronounced, as depth increases (Bennett, 2004). At depths greater than 6.5 ATA, human performance deteriorates significantly, and at depths greater than 10 ATA, signs and symptoms are severe, with the possibility of a diver becoming unconscious (Bennett, 2004).

The general or average threshold depth of 30msw has been established by objective measurements of performance decrements. Subjective reporting of depth of onset of symptoms varies widely, however,

(15)

it is strongly influenced by suggestibility. Recently, computerised measures that are more sensitive have found some indications that psychomotor function may be impaired between 10 and 30msw in a laboratory setting (Petri, 2003). It is not clear whether these impairments have practical implications for real dives.

The signs and symptoms are similar to those seen in alcohol intoxication and the early stages of hypoxia and anaesthesia, with an equally wide variation in susceptibility (Bennett, 2004, p 226). The issue of individual susceptibility is discussed in more detail under Section 2.5 General Findings. It is particularly the euphoria, light-headedness, and motor incoordination of narcosis that has been likened to alcohol intoxication (Lowry, 2002). There is evidence of a significant correlation between alcohol and NN (at 5–6 ATA, breathing air) on both subjective and behavioural measures, which suggest that a biological factor may account for the variations in intensity experienced among individuals (Hobbs, 2008; Monteiro, Hernandez, Figlie, Takahashi & Korukian, 1996). Interestingly, it appears that the degree to which an individual’s performance is impaired by alcohol can predict the degree of impairment when narcotic (Hobbs, 2008).

Studies into the metacognition of divers showed the inaccurate nature of subjective judgements in their own abilities while experiencing narcosis (Harding, Bryson & Perfect, 2004). Metacognition refers to the level of awareness of cognitive functioning, with the implication that, if individuals can sense their own level of cognitive performance accurately (as well as a decrease in performance), then they can react appropriately to that situation. For example, Mount & Milner (cited in Harding et al., 2004) told divers to expect narcosis at different depths, and that is exactly what divers reported during their dives. The implications were that divers were not aware of the true level of their impairment, and in fact, that they tended to over-estimate their ability because of incorrect self-beliefs that they did not suffer from narcosis, despite evidence to the contrary. Harding et al. (2004) found that divers believed their performance at 5 ATA was equal to their performance at 1 ATA on reaction time tasks, and better on a long-term memory task at 5 ATA than at 1 ATA, even when the actual outcome was poorer. They further report that a cohort of divers reported symptoms of narcosis to start at depth deeper than would normally be expected, indicating a possible denial of symptoms, or at least a lack of awareness of symptoms.

2.2.2. Other inert gases besides nitrogen

The same signs and symptoms of NN, or breathing air at more than 4 ATA, have also been observed with other metabolically inactive gases, e.g. neon, argon, krypton, xenon, hydrogen, and the anaesthetic gases, although at different partial pressures (Fowler et al., 1985; Lowry, 2002). The use

(16)

diving at increased depth or duration. The present study is concerned with the effects of narcosis when breathing normal air (i.e. surface mixture) under specific pressures (i.e. 6 ATA), within limited decompression schedules. However, a quick overview of other gases is provided for the sake of comprehensiveness (from Bennett, 1993).

Helium does not lead to narcosis, and is thus used for deep and saturation diving. It has a number of disadvantages, e.g. voice distortion, high thermal conductivity, and cost. It further is associated with High Pressure Nervous Syndrome (HPNS) (at depths of 150m+), although the association may not be an effect of helium, but rather due to the extreme depth, which also requires the use of this gas to reach it. HPNS must be distinguished from IGN (Section 2.2.7). Neon does not lead to narcosis either, but is denser than helium, and less often used due to its possible adverse respiratory effects, i.e. increased work breathing relative to hydrogen. Hydrogen is less narcotic than nitrogen, but dangerous as it becomes explosive in mixtures of more than 4% oxygen. The addition of hydrogen to heliox delays to greater depths the signs and symptoms of HPNS. Argon is 2 to 4 times more narcotic than nitrogen. Xenon too has a much greater narcotic effect than nitrogen, and produces anaesthesia in humans at depths of 10m. Krypton and nitrous oxide are both very narcotic.

2.2.3. Human performance models

A historical overview of the developments around the scientific understanding of IGN can be found in Bennett (1993, 2004) and Lowry (2002). The aetiology of IGN lies in the realm of biophysiology and biochemistry, and is beyond the scope of this thesis. A number of hypotheses have been investigated in the literature, including the lipid solubility hypothesis, the critical volume concept, and the multi-site expansion model. Brief reviews can be found in Bennett (1993) and Lowry (2002), and more recently in Smith & Spiess (2010) and Rostain, Lavoute, Risso, Vallée & Weiss (2011).

The site of action of narcosis in the brain probably lies at the synapses. Mechanisms involving interference with the electrochemical mechanisms necessary for the transfer of electrical potential across synaptic gaps is thought to contribute to narcosis (Bennett, 2004). Polysynaptic regions of the brain, such as the ascending reticular formation system and the cortical mantle, are likely to be regions of the brain most affected (Bennett, 2004). Studies have investigated inhibitory and excitatory neurotransmitters and receptors in the central nervous system (CNS), including noradrenaline, serotonin, dopamine, gamma-amino butyric acid (GABA) and glycine (Lowry, 2002). Exposure of rats to narcosis raised cellular dopamine in the brain areas controlling the extra-pyramidal system, which may account for some of the neuromuscular disturbances of IGN (Barthelemy-Requin, Semelin & Risso, 1994).

(17)

A number of human performance models for IGN have been formulated (cf. Fowler et al., 1985, for an earlier review), including the descriptive model, the hierarchical organisation model, the operant paradigm, and the slowed processing model.

The descriptive model defines NN as a set of signs and symptoms, and identified performance decrements at increased pressure for, among other things, arithmetic (Bennett, 1993; Mears & Cleary, 1980), long-term memory (Fowler, 1973; Philp, Fields, Roberts, 1989; Tetzlaff, Leplow, Deistler, Ramm, Fehm-Wolfsdorf, Warninghoff & Bettinghausen, 1998; Vaernes & Darragh, 1982), reaction time (Bennett, 1993), and motor dexterity (Bennett, 1993; Mears & Cleary, 1980) functions. Motor performance seem more resilient to the effects of narcosis than cognitive functions (Abraini, 1997; Abraini & Joulia, 1992). This conceptualisation has practical value, as accurate prediction of performance will lead to better preparation to avoid or manage these impairments. However, it has limited value for theoretical conceptualisations.

The hierarchical organisation hypothesis posits that ‘higher’ physiological centres are more sensitive to anaesthetics, and that they will be affected first. This implies that the more complex mental functions should be affected more quickly by narcosis than the less complex functions. This model has limited explanatory power (Fowler et al., 1985).

The operant paradigm draws on the classical theories of behaviourism, and it was developed for use in animal studies on the effects of narcosis. This allowed for studies on the comparative effects of inert gases, and their progressive effects at greater pressures. There are suggestions that results with this technique can be generalised to humans, making it a useful avenue for future study (Fowler et al., 1985).

A more sophisticated approach explains narcosis through the slowed information processing model (cf. Fowler et al., 1985 for an earlier review). The effect of narcosis is thus ascribed to a single underlying functional effect: slowing due to decreased arousal. This is manifested by an increase in response time, and often by a decrease in accuracy. To understand how narcosis disrupts complex tasks, one needs to regard the information processing system as being dynamic and as involving factors such as cumulative slowing in working memory and alterations in strategy. The following discussion, unless otherwise indicated, is adapted from Fowler et al. (1985, p.377–387).

(18)

Environmental Response

Input Output

Figure 2.1: Effects of exposure to compressed air (Fowler et al., 1985).

The information processing system has three main features:

The first is a series of stages, each of which takes time to process information. These stages are referred to as structural variables and are named after the kind of processing they perform – perceptual, stimulus-response, and effector stages. Other kinds of structural variables are memory stores, which hold information either temporarily or permanently. Each stage takes time to process information and adds an increment to the total time required for a response.

The second feature is that overall efficiency is determined by the level of arousal or activation – called a functional variable. Up to a certain point, an increase in arousal improves performance (Dodge-Yerkson law), where after it interferes with optimal performance. An important role assigned to the ascending reticular activating system is regulating cortical excitability.

The third feature encompasses strategic variables, and consists of control processes, which organise the resources of the system to accomplish a particular task. These processes give the systems its complex dynamic characteristics and may or may not be under the conscious control of the performer. They include the distribution of attention, decision criteria, rehearsal strategies, and speed-accuracy trade-off settings (e.g. how performers strike a balance between speed and accuracy). The strategy used to perform a task is likely to change under narcosis, but is secondary to slowing, i.e. as a compensatory response to slowing or in reaction to euphoria.

Narcotic performance deficits, reflected in decreases in the speed or accuracy of responses, could be due to structural, functional, or strategic variables, either alone or in combination.

Perceptual Processing Long-term Memory Short-term Working Memory Decision and Control Processing Effector Processing

(19)

Evidence indicates that slowing due to narcosis is not caused by a disruption of processing stages (i.e. it is not caused by structural deficiency), but by a decrease in the efficiency of the system as a whole (i.e. functional changes). Support for this came from studies on the additive effect of alcohol (a CNS depressant) on slowing (Hamilton, Fowler & Porlier, 1989), and the ameliorative effect of amphetamine (a CNS excitant) on slowing (i.e. less slowing) (Hamilton at al., 1989). Further, under conditions of narcosis, time is judged to pass more quickly, which may imply a slowing down of an ‘internal clock’ (i.e. a functional slowing of the CNS).

There is further tentative support that a decrease in accuracy may be due to a strategic rather than structural change – in other words, due to a shift in speed-accuracy trade-off setting rather than a distortion in information transmission. Two reasons for this have been proposed: Firstly, divers distribute the impact of narcosis between speed and accuracy so the loss of speed is not so noticeable, and secondly, the euphoria induced by narcosis makes divers less reluctant to maintain a particular setting. Thus, task completion strategies may change under conditions of narcosis, but such changes are secondary to slowing, i.e. they appear as compensatory responses to slowing or in reaction to euphoria (Fowler, Hendriks & Porlier, 1987; Fowler, Mitchell, Bhatia & Porlier, 1989).

The role of working memory may be critical in understanding the effect of narcosis on complex tasks. It has been hypothesised that working memory processing occurs in a series of operations (Baddeley & Hitch, 1994). If each operation performed is slowed by narcosis, then due to the iterative nature of processing, there will be cumulative slowing. Response time will increase, as the total number of operations that must be performed increases. Support for this hypothesis came from studies involving complex tasks, like arithmetic. This perspective may also provide an explanation of why psychomotor tasks are less sensitive to narcosis than are cognitive or response time functions – dexterity requires fewer mental operations (thus leading to less cumulative slowing).

Subjective symptoms of narcosis (e.g. euphoria, state of consciousness, inhibitory state) may further influence performance, for example, by causing shifts in the value of strategic variables (e.g. a lack of caution may mean that accuracy is allowed to deteriorate), by distracting attention, and by creating anxiety, which in itself can degrade performance.

Currently, there is good provisional evidence in support of the slowed information processing model (Hamilton et al., 1989), and previous weaknesses of the theory in explaining memory loss (Lowry, 2002) have recently been adequately addressed (Hobbs & Kneller, 2009).

(20)

For example, anxiety in itself impairs performance, even at the surface. The water effect refers to the limitations a diver faces when entering the underwater environment, even at shallow depths, due to the issues of movement in water, the use of life-support equipment, and so forth (Weltman, Christianson & Egstrom, 1970), all which need to be isolated from the actual effect of inert gas (narcosis). The water effect is discussed in more detail in Section 2.3.1, and anxiety in Section 2.4.1.

2.2.4. Creating context to review specific findings

The findings from available studies (which will be discussed in Sections 2.2.5 and 2.2.6) need to be interpreted cautiously. It is difficult to study the effects of narcosis systematically, due to the many variables that need to be controlled, and the costs that are involved in large sample studies.

Due to high costs, studies conducted used limited numbers of dives, and/or used working dives to collect their data, where less experimental manipulation was possible. For the same reasons, sample sizes are generally small (cf. Weybrew, 1978, for an older review on sample sizes).

Some of the variables that confound comparisons between studies include:

1) The technical aspects of experimental conditions, e.g. gas mixtures, gas partial pressures, temperature, and rates of compression;

2) The demographic backgrounds of test subjects, e.g. type (professional vs. sport divers vs. non-divers), age of participants, their training and exposure, and motivation to participate; 3) Environmental conditions, e.g. wet vs. dry (Baddeley, 1966; Baddeley, DeFigueredo,

Hawkswell-Curtis & Williams, 1968; Baddeley & Flemming, 1967; Baddeley & Idzikowski, 1985), the use of protective clothing, and the other potentiating factors discussed in the next section;

4) Different measurements (tests and indicators) are used under different definitions of neuropsychological functions. This makes it difficult, at times, to untangle the underlying construct that the outcomes are supposed to measure. The use of self-reports, particularly in older studies, further complicates the issue.

There are also different methods for reporting results. Some investigators report decrements in performance in terms of percentages, which can be related to either normobaric performance scores or the scores of control groups. Others use standard deviations from the mean, which are in turn influenced by the level of interpersonal heterogeneity, which varies across samples and traits. Others still present graphs or charts only.

(21)

Apart from the fact that different studies used different participants and different measurements, and did so under different conditions, inter-individual susceptibility to narcosis further confounds the issue, as different people react differently to narcosis, and are affected by it at different depths. Perhaps because of this, neuropsychological research into the cognitive effects of NN is plagued by inconclusive and sometimes contradictory results.

2.2.5. General findings

Before reviewing the specific neuropsychological effects of narcosis on human behaviour, general findings regarding narcosis and performance will be dealt with, to create a background against which the succeeding section can be interpreted.

The most severe signs and symptoms of narcosis are present immediately when reaching the desired pressure (Bennett, 1993), with narcosis usually more severe immediately on arrival at the specific depth. There may be some improvement shortly afterwards, followed by a stabilising in the level or degree of narcosis (Bennett, 2004). Further ‘improvements’ are likely to be primarily subjective improvements (i.e. some form of mental compensation), as objective tests show no change (see discussion of adaptation that follows later). Signs and symptoms of narcosis are not affected by the duration of exposure.

The signs and symptoms become more marked with greater depth (Bennett, 1993). On decompression, recovery is rapid, with no after-effects, except an occasional amnesia to events while at depth (i.e. during the narcotic state) (Bennett, 2004). Very rapid compression appears to potentiate narcosis (Bennett, 2004). The extent of performance decrement appears directly proportional to task complexity (Kiessling & Maag, 1962).

There is great variation in individual susceptibility to narcosis and the severity of its effects. As mentioned, the threshold for the objective identification of narcotic effects is 30msw. The experience of NN is further highly subjective. One study found that reports of subjective onset of narcosis on average come at 34.5m, with a range of 15 to 60 m (Harding et al., 2004).

The effect of narcosis on performance is greater under wet than dry conditions (Baddeley, 1966; Baddeley & Flemming, 1967; Baddeley & Idzikowski, 1985). Thus, results from dry chambers cannot always be transferred directly to wet conditions. Inside a hyperbaric chamber, the psychological stressors of ocean diving are absent, as are the muscular strain and environmental conditions (buoyancy, temperature) encountered there (De Moja, Reitano & De Marco, 1987). In the

(22)

open-water conditions (Baddeley et al., 1968; Baddeley & Idzikowski, 1985). Open water also elicits performance interference from the water effect, as well as potential dangers (i.e. anxiety). It is generally accepted that the greater effect of NN in the open water, relative to hyperbaric chambers, can largely be attributed to increased anxiety (Hobbs & Kneller, 2011).

There is no evidence that age or formal education play any significant role in moderating the effect of hyperbaric pressure, although it has been suggested that greater intellectual ability protects divers to some extent against the effects of narcosis (Bennett, 1993). There is further no evidence that previous technical exposure (whether formal training or informal contact) play a role in moderating the effects of NN, although it can be hypothesised that formal exposure, through a process of over-learning of principles or practices, may protect against impaired performance on tasks requiring technical reasoning or manual dexterity.

A number of diver characteristics have been found to potentiate or moderate the severity of narcosis: More experienced divers appear to be less affected (Bennett, 1993; De Moja et al., 1987), but as with intellectual ability, there are large differences in the range of individual responses (Behnke, 1984). Apprehension and anxiety (Davis, Osborne, Baddeley & Graham, 1972) exacerbate the effects of narcosis, especially the impairment of psychomotor dexterity (Baddeley & Idzikowski, 1985; Kneller et al., 2012; Mears & Cleary, 1980). The dangers of a hostile environment may further potentiate the effects of narcosis (Baddeley et al., 1968; Baddeley, 1972). Anxiety is discussed in more detail in Section 2.4.1.

A number of physiological and chemical characteristics have also been found to potentiate the severity of narcosis: Any increases in exogenous or endogenous carbon dioxide potentiate narcosis synergistically (Bennett, 2004; Fothergill, Hedges & Morrison, 1991). Thus it is likely to be more severe in the swimming or working diver wearing a breathing apparatus than in the case of a diver in a pressure chamber (Bennett, 2004). Variations in the oxygen percentage in breathing mixtures further affect the degree of narcosis (Bennett, 2004; Frankenhaeuser, Graff-Lonnevig & Hesser, 1963). Hard work and/or fatigue (Bennett, 1993) facilitate the effects of narcosis, although this might be through the mechanisms of hypercapnia or carbon dioxide retention referred to above. Moderate work or exercise may conversely ameliorate the effects of narcosis due to its role in arousing the CNS (cf. Fowler et al., 1985). No gender effect of NN on neuropsychological performance has been found (Jakovljevic, Vidmar & Mekjavic, 2012).

Ethanol consumption depresses the CNS, and exacerbates the effects of narcosis (Fowler et al., 1987; Hamilton et al., 1989). As mentioned, individuals with greater sensitivity to alcohol also display a greater sensitivity to narcosis. In contrast, amphetamine excites the CNS, and ameliorates the effects

(23)

of narcosis (Fowler et al., 1985; Hamilton et al., 1989); however, its effects may be unpredictable (Lowry, 2002). The effects of ethanol and amphetamines are both in line with the view that narcosis depresses the CNS. The effects of different pharmaceutical preparations are reviewed in Bennett (1993). Some protect against narcosis, while others enhance it. There are indications that lithium carbonate may ameliorate the effects of narcosis (Leach, Morris & Johnson, 1988).

Findings regarding adaptation to the effects of NN are generally inconsistent (Fowler et al., 1985). Some studies suggest that frequent exposure may afford some adaptation (Bennett, 1993; Moeller & Chattin, 1975), although continuous exposure may only lead to adaptation after a number of days – 8 or 9 in one study (Coler, Patton & Lampkin, 1971). Acclimatisation is probably a more accurate description, rather than adaptation, as most other studies found little or no behavioural adaptation to NN in response to short repetitive hyperbaric exposures (Hamilton, Laliberté & Fowler, 1995; Moeller, Chattin, Rogers, Laxar & Ryack, 1981; Rogers & Moeller, 1989; Whitaker & Findley, 1977). Subjective adaptation can occur without parallel performance improvement (Hamilton, Laliberté & Heslegrave, 1992), and the anecdotal reports of adaptation by divers can probably be attributed to the subjective rather than the behavioural components of narcosis (Hamilton et al., 1992).

Improvements in performance are generally attributed to non-specific learning, and the adaptation of subjective symptoms is often mistaken for performance adaptation (Hamilton et al., 1995). The importance of non-specific learning to lessen the effects of narcosis emphasise the value of practice and over-learning of behaviour, prior to underwater work.

2.2.6. Clinical manifestations

Attempts to quantify the effects of IGN can be roughly divided into two methods (Lowry, 2002): measuring the behavioural expression on neuropsychological tasks, and observing changes in neurophysiological parameters. The latter lies outside the scope of this thesis and only the neuropsychological (‘behavioural’) expressions of NN will be reviewed here.

It is not always clear how to organise behavioural findings, partly due to methodological issues raised previously, but also because the tests, tasks and measurements used at times cover multiple neuropsychological domains. For descriptive purposes, the clinical manifestations will be organised into three major groups, namely cognitive performance, psychomotor functioning, and mood states and social behaviour. An effort has been made to classify the results of measures according to the major functional activity that they elicit, but it is conceded that, while the allocation presented here tried to follow convention, it was not always possible, and some were assigned slightly arbitrarily.

(24)

In general terms, narcosis has been defined as a slowing in mental activity, with delays in auditory and olfactory stimuli and a tendency to word-idea fixation. The resulting limitation in power of association and perception then becomes dangerous in the presence of overconfidence associated with NN (Bennett, 2004). It is possible to view mental slowing as an – or even the – underlying factor affecting all of the neuropsychological functions below, but for the sake of organisation, reaction time will here be included under cognitive functions.

Generally, higher functions, such as reasoning, judgement, recent memory, learning, concentration and attention are affected first. If the partial pressure of inert gas is further elevated, it results in a progressive deterioration of psychomotor functioning and mental performance, and an increase in automatisms, idea fixation, hallucinations, and culminating finally in stupor and coma (Lowry, 2002). Perceptual narrowing can occur, where divers may become less aware of potentially significant stimuli outside their prescribed task (Lowry, 2002).

A summary of the behaviour characteristics of NN according to depth can be found in Table 2.1 (from Edmonds, Lowry, Pennefather & Walker, 2002, p 186).

Table 2.1: Effects of exposure to compressed air at increased pressure and depth. Pressure (ATA) Effects

2–4 Mild impairment of performance on unpractised tasks Mild Euphoria

4 Reasoning and immediate memory affected more than motor coordination and choice reactions

Delayed response to visual and auditory stimuli

4–6 Laughter and loquacity may be overcome by self-control Idea fixation, perceptual narrowing, and over-confidence Calculation errors; memory impairment

6 Sleepiness; illusions; impaired judgement

6–8 Convivial group atmosphere; may be terror reaction in some; talkative; Dizziness reported occasionally

Uncontrolled laughter approaching hysteria in some 8 Severe impairment of intellectual performance

Manual dexterity less affected 8–10 Gross delay in response to stimuli

Diminished concentration; mental confusion

10 Stupefaction

Severe impairment of practical activity and judgement Mental abnormalities and memory deficits

Deterioration in handwriting; uncontrollable euphoria; hyperexcitability; Almost total loss of intellectual and perceptive faculties

>10 Hallucinogenic experiences Unconsciousness

(25)

Cognitive performance

Cognitive functions are affected by narcosis to a greater extent than neuromuscular functions.

Attention and concentration

Choice reaction time is consistently impaired under narcosis, indicating a slower processing of information under pressure (Bennett, 1993; Fowler et al., 1989). In general, more time is needed to complete tasks. Table 2.2 presents a sample of the choice reaction time studies done to date.

Concentration, as measured by a modified Stroop Test at 6 ATA breathing air, showed an 11% decrement in scores (Fothergill et al., 1991). Digit-symbol substitution (WAIS-R) at 30 m (4 ATA) breathing air did not show any difference from a control group (Gallway, Millington, Wolcot, Mirsky, Van Gorp & Wilmeth, 1990), although the depth may not have been enough to elicit measurable effects.

Visual scanning seems particularly sensitive to the effect of cumulative depth. One study found that performance on the Trailmaking Test (A&B) at 4 ATA, breathing air, did not differ from a control group (Gallway et al., 1990). The same test administered at 7 ATA, breathing air, found significant impairment (Van Wijk, 2008a). Other studies also reported decrements in visual scanning performance at 7 ATA (Moeller et al., 1981; Whitaker & Findley, 1977), although the extent of the decrement was not reported. Table 2.3 presents a sample of the visual scanning studies done to date. It was hypothesised that the pressure at 4 ATA may not have been enough to elicit clear decrements in performance, and that the impairment of performance at 7 ATA could then be attributed to the observation that increased pressure increases the impairment of cognitive functions (Bennett, 1993).

Time estimation is consistently affected, with divers giving significantly longer time estimations under pressure. There are suggestions that this may possibly be due to the water effect as much as to narcosis (Mears & Cleary, 1980), or that it may refer to the slowing of the internal clock (Fowler et al., 1985). Table 2.4 presents a sample of the time estimation studies done to date.

(26)

Table 2.2: Studies investigating choice reaction time.

Conditions Outcome Source Pressure Breathing gas

100 fsw (±4 ATA) Air 21% decrement in response time Kiessling & Maag, 1962 188 fsw (±7 ATA) Air Significant slowing in response

latency

Whitaker & Findley, 1977

6 ATA Air No effect Abraini & Joulia, 1992

7 ATA Air Small but significant decrement

of 5%

Abraini & Joulia, 1992

Table 2.3: Studies investigating visual scanning and tracking.

30msw (4 ATA) Air No effect Gallway et al., 1990

188fsw (±7 ATA) Air Significant impairment in pursuit tracking task

Whitaker & Findley, 1977

7 ATA Unknown Decrement in visual scanning

performance

Moeller et al., 1981

7 ATA Air Trail A: 46% decrement

Trail B: 54% decrement (completion time)

Van Wijk, 2008a

Table 2.4: Studies investigating time estimation.

106fsw (±3.5 ATA)

Normoxic-nitrogen mixture

Significantly longer time estimations

Miller, Bachrach & Walsh, 1976 30msw (4 ATA) Air Significantly longer time

estimations

Mears & Cleary, 1980 122msw (±13

ATA)

Unknown Disorganised sense of time Adolfson, cited in Bennett, 1993

Language

It is not clear to what extent verbal comprehension is affected under pressure. Divers can recognise language as communication, and the ability to comprehend instructions is intact to great depth. At 13 ATA, orders are appreciated but ignored (Adolfson, cited in Bennett, 1993). This is probably because the euphoria interferes with task execution, rather than it suggesting a language deficit.

(27)

Memory

No previous studies with hyperbaric pressure and tactile memory could be located.

The impairment of verbal memory (as assessed by word lists) has been well documented (cf. Hobbs & Kneller, 2009; Tetzlaff et al., 1998), especially the storage of information in long-term memory (LTM). Short-term memory does not seem to be affected by narcosis to the same degree (Fowler, 1973; Fowler et al., 1987; Morrison & Zander, 2008; Philp et al., 1989). Verbal memory, encoded under conditions of hyperbaric pressure, is impaired when free recall is tested under both hyperbaric and normobaric conditions (Hobbs & Kneller, 2009; Vaernes, 1982; Vaernes & Darragh, 1982; Tetzlaff et al., 1998). It has been suggested that the effects of narcosis on memory could be explained by a disruption of the processing that is required to both encode information into long-term memory and to retrieve it (Tetzlaff et al., 1998). A biological factor might also be implicated, as a reduction in recall capacity under conditions of narcosis is associated with both prolactin and testosterone levels (Vaernes & Darragh, 1982).

However, cued recall appears to correct the deficit in free recall (Fowler, White, Wright & Ackles, 1980; Gallway et al., 1990; Hobbs & Kneller, 2009; Philp et al., 1989; Tetzlaff et al., 1998). Cueing refers to the process enacted by an individual to retrieve information through the use of external memory cues (Hobbs & Kneller, 2009). In other words, hyperbaric pressure seems to have less effect on memory recognition than free recall. This suggests that information is stored in LTM (i.e. the process of encoding is not disrupted), but that retrieval of learnt information (under narcosis) is impaired.

Then again, as recall at surface after a dive (of material learned under narcosis) is also affected, an impairment of self-guided search is an inadequate explanation of the effect of narcosis on memory. A recent hypothesis (Hobbs & Kneller, 2009) postulates that the pattern of impairment reported occurs because encoding under narcosis produces a weaker memory trace than normal. According to the information processing model, narcosis disrupts processing when learned material is encoded (Fowler et al., 1985), resulting in fewer cognitive resources being available for encoding. Thus the material is learned, but the quality of encoding and processing is reduced, producing a weaker memory trace. Hobbs & Kneller (2009) provide some theoretical support for this. In comparison, when possible responses (‘cues’) in a recognition task are provided, a less ‘sophisticated’ level of processing is required. This explanation would explain why material learned under narcosis would be harder to recall than material learned in shallow water, regardless of whether the act of recall took place in shallow or deep water.

(28)

This is supported by findings that, in general, over-learned tasks are not affected, but new learning is negatively affected by narcosis (Fowler et al., 1987; Vaernes, 1982; Vaernes & Darragh, 1982), possibly because over-learned tasks were originally encoded under normobaric conditions. Further support comes from a study that found that rehearsal strategies changed under narcosis. It was hypothesised that the reduction in efficiency of encoding could be caused by the slowing of information processing, in that fewer words are rehearsed as a strategy to encode the remaining words more effectively (i.e. if there is slower processing, then there is a need to compensate for this loss of efficiency – by rehearsing fewer words) (Fowler et al., 1987).

Current evidence indicates that memory decrements in hyperbaric chambers are comparable to those underwater (Hobbs & Kneller, 2009). Cueing is thus of practical value, as it may be an important strategy for eliciting information acquired during exposure to narcosis (Philp et al., 1989). Free recall from the primacy2 regions is most impaired by hyperbaric pressure, with little effect on free recall from the recency region (Fields, 1986, cited in Fowler et al., 1987; Philp et al., 1989). Recall of earlier information may thus be at particular risk if encoded under narcosis.

Verbal memory, if assessed through logical memory (using the WMS-R) and learned at 4 ATA, shows no impairment (Gallway et al., 1990). Logical memory has a cued component, which might explain the negative finding. Table 2.5 presents a sample of the verbal memory studies done to date.

Table 2.5: Some studies investigating verbal memory.

36msw (±4.5 ATA)

Air 20% decrement in immediate

recall from primacy regions

Philp et al., 1989 50% decrement in delayed recall

recognition unaffected 50 msw (6 ATA) Air recall of material learned at

surface not affected when recalled at depth

Tetzlaff et al., 1998

significant impairment of recall at depth of material learned under narcosis

significant impairment of recall at surface of material learned under narcosis

Recognition not significantly affected

(29)

40 msw (5 ATA) Air 10% loss at immediate recall Morrison & Zander, 2008

18% loss after 5 min recall 39% loss after 2 hour delay (22% loss of basic info, 51% loss of abstract info)

37–40 msw (±5 ATA)

Free recall of material learned at surface not affected when recalled at depth

Hobbs & Kneller, 2009

Significant impairment of recall at depth of material learned under narcosis

Significant impairment of recall at surface of material learned under narcosis

Time delay before recall is important. In other words, information encoded at depth degrades quickly, and longer delays result in greater loss of information (Morrison & Zander, 2008). Further, more complex or abstract information is lost more quickly than simple information (at depth and on surface), and this too increases with time delay (Morrison & Zander, 2008). As information that has meaning is easier to remember than abstract information (Baddeley, 1966), the quick decay of abstract information (learned and retrieved at depth) may possibly be due to a more ‘shallow’ encoding of such abstract information (Morrison & Zander, 2008).

Conceptual functions

Studies on the effect of narcosis on conceptual reasoning provide inconsistent results, mainly because of the use of different tests to measure it. In general, reasoning is more impaired with greater depth (Bennett, 1993), and with more unfamiliar material. Table 2.6 presents a sample of the studies done on reasoning to date. Other discussions (Bennett, 1993; Baddeley, 2004) also report that reasoning is impaired, but do not report the extent of such impairment at increased depth. Decrement on reasoning tests has been correlated to cortisol levels at 7 ATA. Higher cortisol levels were associated with better performance (Vaernes & Darragh, 1982), which seems to support the position of the slowed information processing hypothesis that stimulation of the CNS ameliorates the effects of narcosis.

Errors in the recording of arithmetic data have also been reported (Bennett, 2004), but this seems to be a problem of perception rather than mathematical reasoning.

(30)

Table 2.6: Studies investigating reasoning.

Pressure Type of reasoning Outcome Source

Unknown Arithmetic

(mathematical reasoning)

Decreased under pressure Shilling & Willgrube, 1937

61msw (±7 ATA) Arithmetic task Decreased from 20 to 15 points on task

Adolfson, cited in Bennett, 1993 76msw (±8.5

ATA)

Errors increased from 6 to 22 (increase of 250%)

122msw (±13 ATA)

61% less questions correctly answered, with 25% more errors

6 ATA Arithmetic task 22% decrement in scores (combination of answers attempted, answered correctly, and errors made)

Fothergill et al., 1991

100fsw (±4 ATA) Conceptual reasoning test

33% decrement in time to solve problem

Kiessling & Maag, 1962

4 ATA Visual reasoning

(like matrices)

20% impairment in correct answers within time limit

Synodinos, 1976 But prior practice in shallow

water cancels effect totally 30msw (4 ATA) Conceptual

reasoning (progressive matrices)

26% decrement in

performance during day dive

Mears & Cleary, 1980 36% decrement in

performance during night dive

60msw (7 ATA) Reasoning test No significant reduction Vaernes, 1982 60msw (7 ATA) Reasoning test Significant decrease (when

processing unfamiliar material)

Vaernes & Darragh, 1982

7 ATA Number ordination

task

14% decrement in performance

Abraini & Joulia, 1992 16% decrement in

performance

Abraini, 1997

7 ATA WISC Mazes 9% decrement in performance

(compared to 1 ATA)

Van Wijk, 2008a But possible neuromuscular

interference Picture completion

task (WAIS)

8% decrement in performance (compare to 1 ATA)

(31)

Psychomotor functioning

Studies on the effect of NN on manual dexterity and motor coordination remain inconclusive. In 1962, Kiessling & Maag found significant impairment at 4 ATA for complex tasks, but since then no evidence has been found that mild states of NN lead to significant psychomotor impairment on simple tasks at pressures within sport diving limits (Abraini, 1997; Abraini & Joulia, 1992; Biersner, Hall, Linaweaver & Neuman, 1978; Gallway et al., 1990). In fact, dexterity appears the neuropsychological function least affect by mild hyperbaric pressure (Abraini & Joulia, 1992; Abraini, 1997; Biersner et al., 1978; Fowler et al., 1985). However, decreased manual dexterity at depth has been reported (see Table 2.7), which potentially could reduce a diver’s ability to operate equipment and increase the risk for mistakes during emergencies (Kneller et al., 2012).

Handwriting

Handwriting becomes larger, as narcosis becomes more severe (Bennett, 2004). In a digit-copy task, at 4 ATA breathing air, handwriting became 21% larger in size. This was not affected by practice (handwriting is over-practiced behaviour) (Synodinos, 1976).

Fine visual-motor coordination and finger dexterity

Fine motor manipulation is impaired under pressure, usually from over-exaggeration of movements (Bennett, 2004). If movements are carried out more slowly than usual, the impairment of efficiency is likely to be less severe (Bennett, 2004). Simple motor dexterity is less affected, and complex motor dexterity more affected (often through requiring more time to complete the task, in line with the cumulative slowing of the slowed information processing model). When the Pin Test was administered in water at 4 ATA, a 17% decrement in perceptual-motor skill was found (De Moja et al., 1987). Table 2.7 presents a sample of the fine visual-motor coordination studies done to date.

Gross psychomotor skills

Gross manual skills are more resilient than fine manual skills, but also decrease as depth increases (Bennett, 1993).

Balance

Balance is affected under pressure (in a chamber), and postural difficulties increase as depth increases (Bennett, 1993).

(32)

Table 2.7: Studies investigating fine visual-motor coordination.

Pressure Task Outcome Source

76msw (±8.5 ATA)

Manual dexterity Unaffected Adolfson, cited in

Bennett, 1993 91msw (±10 ATA) Practical ability Marked impairment observed

122msw (±13 ATA)

Manual dexterity Reduced by 35% 100fsw (±4 ATA) Modified Purdue

Pegboard

8% decrement in number of parts assembled in 30 seconds

Kiessling & Maag, 1962 5 ATA Mirror drawing Only slight depressant effect

on psychomotor performance

Frankenhaeuser et al., 1963

170fsw (±7 ATA) Ring and Peg Test Simple motor dexterity not significantly affected

Biersner et al., 1978 Complex motor dexterity

significantly affected

30msw (4 ATA) Screw-plate test 45% decrement Mears & Cleary, 1980 36msw (±4.5

ATA)

Ball-bearing test No effect Philp et al., 1989 30msw (4 ATA) Grooved Pegboard No effect Gallway et al., (1990) 6 ATA Purdue Pegboard 9% decrease in parts

assembled in 1 min

Fothergill et al., 1991

6 ATA Manual dexterity

test

No effect Abraini & Joulia, 1992

7 ATA Manual dexterity

test

Small but significant decrements (4%)

Abraini & Joulia, 1992

5 ATA Digit-letter

substitution test

Significant decrement (exacerbated in cases of high anxiety)

Hobbs & Kneller, 2011

4.5–5 ATA Novel manual

dexterity task

12.5% decrease compared to performance in shallow water (<2 ATA) (due to longer time required)

Kneller et al., 2012

Physical sensation and appearance

Impairments of tactile sensation are highly subjective, and include numbness (decreased sensation) of the skin and a tingling sensation in the lips, legs, and feet at great depths (Bennett, 2004). There is no evidence to suggest that tactile recognition is affected. At extreme depths, there is a characteristic deadpan look to the face (Bennett, 2004).

Mood states and social behaviour

The more ‘dynamic’ aspects of psychological experience of narcosis are a neglected field of research, and only a few studies have examined it in any systematic manner. Most reports of changes in mood, social behaviour, or personality are anecdotal, and it is unclear how such changes were measured.

(33)

Reports further use different understandings of the terms ‘mood’ or ‘personality’, making interpretation difficult. A few studies reported on these changes accompanying very deep saturation diving, usually while breathing helium-oxygen mixtures, and will thus not be included in this review.

Mood states

At depths below 10 ATA, manic and depressive states may occur, and changes in personality have been reported (Bennett, 2004). At shallower depths (between 4 and 10 ATA), divers may experience feelings of well-being and stimulation similar to the overconfidence of mild alcohol intoxication. Occasionally, terror develops, but this is most likely in a novice diver who is apprehensive in a new environment (Lowry, 2002). Laughter, loquacity, light-headed sensations, feelings of stimulation and excitement might be present. Increased effort at self-control may overcome such behaviour to some extent (Bennett, 2004).

A study by Fowler & Ackles (1972) used rating scales to identify four clusters of adjectives to describe the subjective experience of narcosis, namely:

 euphoria (more carefree, cheerful)  state of consciousness (more fuzzy, hazy)  work capability (less effective)

 disinhibitory state (less cautious and self-controlled)

Other studies used self-developed adjective lists to elicit subjective experiences of narcosis (Hamilton et al., 1992, 1995), or more formal adjective-based scales, e.g. modified Subjective High Assessment Scale (SHAS) (Hobbs, 2008; Monteiro et al., 1996).

State of consciousness

There are reports from divers of feeling ‘fuzzy’, and experiencing their environment as ‘hazy’ (Hamilton et al., 1995). They describe a slight sense of disconnectedness. At greater depths (90 m+), unconsciousness can occur (Bennett, 2004).

Social and behavioural disinhibition

Divers may become giggly, loud, and boisterous at depth, and display verbosity and uncontrollable laughing, and so forth (Lowry, 2002). Others may become aggressive, irritated, and so forth, apparently mirroring experiences of alcohol intoxication (cited in Hobbs, 2008). As noted previously, at 122 m, orders were appreciated but ignored (Adolfson, cited in Bennett, 1993). Divers may become less cautious and self-controlled (exhibiting symptoms of ‘recklessness’).

(34)

Emotional processing

It has been hypothesised that the Slowed Processing Model may affect emotional processing similarly to information processing (Löfdahl, Andersson & Bennett, 2013). Löfdahl et al. (2013) found that narcosis at 5 ATA was not enough to impair emotional perception or divers’ ability to determine pleasantness or unpleasantness of different stimuli. However, their results suggested the possibility of lower arousal for unpleasant stimuli. If this can be confirmed, it may indicate a risk of inappropriate or inadequate reactions to danger in threatening situations (Löfdahl et al., 2013).

2.2.7 High Pressure Nervous Syndrome

HPNS is a condition distinct from the IGN that typically occurs at depths beyond 30 ATA, usually breathing oxygen-helium or other gas mixtures. Signs and symptoms include postural and intention tremors, myoclonia, psycho-sensorimotor disturbances, electro-encephalographic changes, and sleep disorders (Bennett, 2004). Mild effects can be found at pressures from 11 ATA, become marked at 19 ATA, and may be debilitating at over 30 ATA (Lowry, 2002). It has been postulated that this syndrome is not due to gas exposure per se, but rather due to a direct effect of pressure on the pre- and post-synaptic membranes (Lowry, 2002).

In contrast to narcosis, there is a greater decrement in psychomotor tests than in intellectual tasks in HPNS. This is due to the associated tremor of hands and arms. HPNS appears to reflect a general excitation of the brain, compared to the decreased excitation seen in IGN (Bennett, 2004).

Overall, divers’ mental abilities are sometimes reduced during high-pressure exposures and are sometimes unaffected (cf. Abraini, Martinez, Lemaire, Bisson, Juan De Mendoza & Therme, 1997; Carter, 1979). However, comparisons between studies are difficult, due to the wide variation between experimental conditions (specifically with regard to depth and gas mixtures employed). For example, perceptual speed was alternatively found to be unaffected (Brady, cited in Carter, 1979), or affected to various degrees (Baddeley, cited in Carter, 1979; Carter, 1979; O'Reilly, 1977), at various depths (ranging from 300m to 600+m) , with possibly greater impairment at greater depths (Logue, Schmitt, Rogers & Strong, 1986).

2.3. Human performance in water

Apart from the effects of pressure and narcosis on human behaviour, the underwater environment also exerts its influence. Performance in water is different from performance on land, as a result of the interaction between human and environmental factors. A brief review is included here for the sake of comprehensiveness.

A cross-over study investigating specific aspects of neuropsychological performance in hyperbaric environments